Treatment of upper limb paresis by transcutaneous electrical nerve stimulation and task-related training during chronic stroke

Abstract

This thesis consists of 2 inter-related studies. The objectives of study 1 were (1) to delineate the characteristics of spasticity, associated reactions, muscle strength, reaction time, and functional performance of the paretic upper limb in patients with chronic stroke, (2) to determine the extent to which they differed from normal subjects, and (3) to delineate the relationships among quantitative variables and clinical assessments of the motor deficits in the paretic UL. The global aim of the main study (study 2) was to examine the effectiveness of a program combining TENS with task-related training (TRT) in promoting motor recovery in the upper extremity of patients with chronic stroke. Ninety-eight stroke survivors participated in study 1. Twenty normal subjects also participated. The quantitative measures included maximum isometric voluntary contraction (MIVC) force of elbow flexors, extensors and hand grip of the affected arm, EMG co-contraction ratios during MIVC of elbow flexors and extensors, associated reactions in the paretic elbow muscles recorded as IEMG during non-paretic hand grip, and reaction time of the paretic wrist in extension and flexion. The clinical assessments included Composite Spasticity Index (CSI), Associated Reaction Rating Score (ARRS), and Wolf Motor Functional Test (WMFT). In the clinical CSI, ARRS and WFMT tests, the intraclass correlation coefficients (ICCs) were very high, with 0.978 for CSI, 0.912 for ARRS, and 0.987 for functional ability and 0.872 for time score of the WFMT, and P values <0.001 for all. The quantitative variables including MIVC force, IEMG and RTs also showed relatively high ICCs ranging from 0.802 to 0.928. The ICCs for MIVC of the elbow extensors and flexors and for hand grip force ranged from 0.804 to 0.863. The ICCs for the IEMG ranged from 0.802 to 0.928. The reaction time for wrist flexion and extension in patients with stroke ranged from 0.863 to 0.883. These results all had P values <0.001. Our findings showed that the affected UL in patients with stroke produced significantly smaller force during MIVC of elbow flexors, extensors and hand grip than those of their non-affected UL and of normal subjects (P<0.01). There were no significant differences in the co-contraction ratio of maximum isometric voluntary(MIV) elbow flexion and extension among the affected and unaffected sides of the stroke survivors, and the normal subjects. When the stroke survivors performed a maximum grip using their non-paretic hand, associated reaction was manifested as elbow flexion (62.2 %), elbow extension (27.6%), or no elbow movement (11.2 %) in the paretic arm. Reaction time (RT) of wrist flexion and extension in the stroke survivors' affected hands were significantly longer than that in the normal subjects and their unaffected hands (P<0.01 for both). Statistically significant correlations were found between MIVC force recorded during elbow flexion in the affected arm and ARRS (negatively; p=-0.321, P=0.001), and WMFT functional ability (p=0.380, P<0.001) and time score (negatively; p=-0.389, P<0.001). MIVC force recorded during elbow extension in the affected arm was found to produce similar results. It correlated with ARRS (negatively; p=-0.291, P=0.004), and with WMFT functional ability (p=0.277, P=0.006) and time score (negatively; p=-0.302, P=0.002). Maximum hand grip force in the affected arm correlated moderately with CSI (negatively; p= -0.425, P<0.001), ARRS (negatively; p=-0.430, P<0.001), and with WMFT functional ability (p=0.658, P<0.001) and time score (negatively; p=-0.630, P<0.001). There were no significant associations between the co-contraction ratios during MIV elbow flexion and the CSI, ARRS, and WMFT results. However, the co-contraction ratio during MIV elbow extension correlated moderately but significantly with CSI (p=0.227, P<0.05), ARRS (p=0.377, P<0.001), and with WMFT functional ability (negatively; p=-0.358, P<0.001) and time score (p=0.360, P<0.001). Moderate but statistically significant correlations were also found between the paretic biceps IEMG recorded as an index of associated reaction during non-paretic hand grip and CSI (p= 0.418, P<0.001), ARRS (p=0.557, P<0.001), and with WMFT functional ability (negatively; p=-0.561, P<0.001) and time score (p=0.559, P<0.001). Although the paretic triceps IEMG recorded as an index of associated reaction during non-paretic hand grip correlated marginally with CSI (p=0.199, P=0.05); like the biceps IEMG, it correlated moderately with ARRS (p=0.371, P=0.001), and with WMFT functional ability (negatively; p=-0.378, P<0.001) and time score (p=0.403, P<0.001). The wrist flexion RT correlated moderately with CSI (p=0.412, P<0.001) and ARRS (p=0.341, P<0.001), and with WMFT functional ability (negatively; p=-0.531, P<0.001) and time score (p=0.504, P<0.001). Similarly the wrist extension RT correlated moderately with CSI (p=0.429, P<0.001), ARRS (p=0.374, P<0.001), and with WMFT functional ability (negatively; p=-0.531, P<0.001) and time score (p=0.486, P<0.001). In summary, our findings from study 1 showed that all the quantitative measures and clinical assessments were reliable, with ICCs ranging from 0.802 to 0.987. Moreover, MIVC force of the affected elbow flexors, extensors and hand grip in patients with chronic stroke was significantly smaller, and RT of their wrist flexion and extension was significantly longer than those of their non-affected UL and of normal subjects. During non-affected hand grip, associated reaction was mainly manifested as elbow flexion (62.2%) in the paretic UL. These 3 quantitative parameters were further found to be correlated moderately but significantly with the clinical scales of CSI (except for MIVC force of elbow flexors and extensors), ARRS, and WMFT functional ability and time scores, in either a positive or negative manner. These findings suggest that both quantitative and clinical assessments could serve as reliable and valid assessment tools to measure treatment effectiveness in patients with stroke over time in study 2. The research design of study 2 was a randomized, controlled trial involving 77 subjects being randomly allocated to 4 groups. One group received TENS alone (n=20), another p-TENS + TRT (n=20), a third received TENS + TRT (n=18), and there was also a control group which received no active treatment (n=19). Outcome measures were recorded in the paretic arm as follows: (1) the composite spasticity index (CSI), (2) maximum isometric voluntary contraction (MIVC) force of elbow flexors and extensors, and hand grip, (3) reaction times (RT) of wrist flexion and extension, and (4) functional ability and time scores of the Wolf Motor Function Test (WMFT). Assessments were carried out before treatment on day 1 (baseline assessment), at week 4 (mid-way through the treatment), at the end of the 8-week treatment program, and at follow-up 4 weeks after treatment ended. Significant differences between groups were found in time domains but not muscle change strength in the UL. After 8 weeks of treatment, the TENS+TRT group showed a significantly greater percentage decrease in the reaction time of wrist flexion (-16.8%) when compared with the TENS group (22.5%, P<0.05) and the control group (26.5%, P<0.05), and the p-TENS+TRT group had a significantly greater percentage decrease in the reaction time of wrist extension (-12.1%) when compared with the TENS group (19.3%, P<0.05%). However, at follow-up 4 weeks after treatment ended, only the TENS+TRT group had a significantly greater percentage decrease in wrist flexion RT (-11.6%) when compared with the control group (31.1%, P<0.05), and of wrist extension RT (-11.5%) when compared with the TENS group (26.5%, P<0.05). With regard to WMFT, the 2 groups receiving TRT (i.e. p-TENS+TRT and TENS+TRT) showed a significant percentage decrease of the WMFT time when compared with the control group after 8 weeks of treatment. In conclusion, our findings from study 2 showed that the 2 groups receiving TRT had significantly faster RT for either wrist flexion or extension, and faster WMFT time scores after 8 weeks of intervention. In the TENS+TRT group, the percentage decrease of wrist flexion RT compared with the control group and of wrist extension RT compared with the TENS group can even be carried over to the follow-up at week 12. The faster RT and motor functional performance plus the presence of carry over effects in the combined treatment group suggest that combing TRT with TENS would be superior to TENS alone, or no active treatment in patients with chronic stroke.